On Mon, Oct 24, 2022 at 02:33:37PM -0700, Darrick J. Wong wrote: > From: Darrick J. Wong <djwong@xxxxxxxxxx> > > If we're in the middle of a deferred refcount operation and decide to > roll the transaction to avoid overflowing the transaction space, we need > to check the new agbno/aglen parameters that we're about to record in > the new intent. Specifically, we need to check that the new extent is > completely within the filesystem, and that continuation does not put us > into a different AG. Huh. Why would they not be withing the filesystem or AG, given that the current transaction should only be operating on an extent within the current filesystem/AG? IIUC, this is code intended to catch the sort of refcount irec startblock corruption that the series fixes, right? If so, shouldn't it be first in the patch series, not last? -Dave. -- Dave Chinner david@xxxxxxxxxxxxx
